Tuesday, October 25, 2011

Stanford AI MOOC Reflection - Homework 2

I'm definitely not as confident in my success on homework #2 as I was on homework #1.  The probability equations were challenging.  I guess I'm more interested in observing learner behavior.  The course confirms that most students do procrastinate.  I, like many others, popped in at the last minute on Monday to find that we had a 24 hour extension on homework.  Server crash.  Apparently, too many people trying to finish up in the final hours.  Even more interesting was doing a quick Internet search on Bayes Rule to find the first three hits were also churning.  Guess a lot of folks were searching for content support.  Still no discussion forum on the course site, but I'm getting a daily digest from http://www.reddit.com/r/aiclass .  My goal is to do a better job keeping up this week.  Time for a coffee break!

Wednesday, October 19, 2011

Sometimes social learning isn't the best option. Stanford AI MOOC Reflection - Unit Three

I earned a respectable grade on my first homework assignment.  Now, on to unit three.  I've already been warned in the introduction about the difficulty of the probability unit.  I'm not sure if warnings are an effective pedagogical tool for everyone, but I'm definitely paying closer attention.  All I can say about Bayes Rule is too many steps.
While I really want to connect with other students in the class, I haven't had much luck with the forums because I can't really trust the accuracy of the posts I'm reading.  With these probability challenges, I feel like I might get a better result visiting Khan Academy.  Sometimes the social network isn't the smartest choice.  Frankly, there's too much noise which only increases confusion and stress.

Wednesday, October 12, 2011

Stanford AI MOOC Reflection - Unit Two

The first homework assignment was posted today.  I had to finish up state spaces, sliding blocks puzzle, and problems with search before attempting homework.  Apparently, my homework answers are saved, but I can go back anytime before the due date to change answers.  I feel about 80% confident in my responses.  I think I'll sleep on them a few days, read the text, review prior lessons and take a final shot at it on Saturday.  In the meantime, I hope to explore the reddit forums and the #AICLASS hashtag on Twitter to connect with other students.

Tuesday, October 11, 2011

Stanford AI MOOC Reflection - Unit One

I started Stanford's Open AI Course today.  I plan to share reflections on learning in a Massively Open Online Course with over 100,000 other registrants.  We'll see how long they (and I) can hang in there.  I'm already a day behind, as the course began yesterday.  At the end of unit one, I've had an introduction to intelligent agents with vocabulary words such as deterministics, stochastic, discrete, continuous, benign, and adversarial.  A machine translation exercise using a Chinese/English menu as an example gave me a new appreciation of machine learning.  I answered 62% of my unit one quiz questions correctly, but learned from my mistakes.  Hopefully, I'll  do better than that on the tests that count.
The discussion feature is not yet working.  I look forward to the reflections of others in the course.  At the moment, it's a lonely endeavor, but I am learning.

Here you see unit 1, Welcome to AI, in which the video lecture runs on the right with progress checked off on the left.  I can check my progress on quizzes with the Progress tab across the top.



Unit 2, Problem Solving, was a bit more challenging.  I thought I understood breadth first search, but my quiz answers beg to differ.  I will go over it again, but this is a point of frustration where I would like to ask questions and/or challenge the instructor.  I'm thinking about how I would approach the learning if sitting in a classroom rather than here online.  I definitely feel the pressure to go back and review.  If sitting in the classroom, I doubt I would ask the professor to repeat himself as many times as I've gone back and reviewed the video.

Evernote is turning out to be a great learning tool.  I can take notes on the lectures and integrate screen captures throughout.  It syncs to my iPad and iPhone, as well.  Here's a shot of my full desk top, Evernote on the left with screen caps from the course and lecture on the right.  Better than being there.